meegsim.configuration.SourceConfiguration#

class meegsim.configuration.SourceConfiguration(src, sfreq, duration, random_state=None)[source]#

This class describes a simulated configuration of sources of brain activity and noise.

Attributes:

srcSourceSpaces: Source spaces object that stores all candidate source locations.
sfreqfloat: Sampling frequency of the simulated data, in Hz.
durationfloat: Length of the simulated data, in seconds.
random_stateint or None, optional: Random state that was used to generate the configuration.

Methods

`plot`(subject[, hemi, colors, scale_factors, ...])	Plot the source configuration.
`to_raw`(fwd, info[, sensor_noise_level])	Project the activity of all simulated sources to sensor space.
`to_stc`()	Obtain an `stc` object that contains data from all sources in the configuration.

__getitem__(name)[source]#

This function provides quick access to the simulated sources. The syntax is similar to list/dict access: sc[name] returns the corresponding source if it is present in the configuration.

Parameters:

namestr: Name of the source.

Returns:

sourcePointSource or PatchSource: The corresponding source.

Raises:

ValueError: In case there is no source with the provided name.

plot(subject, hemi='lh', colors=None, scale_factors=None, show_noise_sources=True, show_candidate_locations=False, **plot_kwargs)[source]#

Plot the source configuration. This function is built on top of mne.SourceEstimate.plot(), and the meaning of all parameters is kept the same unless mentioned.

Parameters:

subjectstr: The name of the subject.
hemistr: The hemisphere to plot. The values “both” and “split” are also supported.
colorsdict or None, optional: This dictionary can be used to override the default color palette, which is described in the Notes section below.
scale_factorsdict or None, optional: This dictionary can be used to override the default display size for sources of different types. See Notes for the default values.
show_noise_sourcesbool, optional: Controls whether noise sources should be displayed (True by default).
show_candidate_locationsbool, optional: Controls whether candidate source locations (all vertices present in the source space) should be displayed (False by default).
**kwargs: dict, optional: Additional parameters that should be passed to mne.SourceEstimate.plot().

Warning

Plotting all candidate locations might require lots of resources depending on the size of the source space.

Returns:

figuremne.viz.Brain or matplotlib.figure.Figure: The resulting figure.

Notes

The following parameters are used for plotting by default:

DEFAULT_COLORS = dict(
    point="green",
    patch="Oranges",
    noise="black",
    candidate="yellow"
)

DEFAULT_SCALE_FACTORS = dict(
    point=0.75,
    noise=0.3,
    candidate=0.05
)

DEFAULT_PLOT_KWARGS = dict(
    background="w",
    cortex="low_contrast",
    colorbar=False,
    clim=dict(kind="value", lims=[0, 0.5, 1]),
    transparent=True,
)

to_raw(fwd, info, sensor_noise_level=None)[source]#

Project the activity of all simulated sources to sensor space.

Parameters:

fwdForward: The forward model.
infoInfo: The info structure that describes the channel layout.
sensor_noise_levelfloat, optional: The desired level of sensor-space noise between 0 and 1. For example, if 0.1 is specified, 10% of total sensor-space power will stem from white noise with an identity covariance matrix, while the remaining 90% of power will be explained by source activity projected to sensor space. By default, no sensor space noise is added. See Notes for more details.

Returns:

rawRaw: The simulated sensor space data.

Notes

The adjustment of sensor space noise is performed as follows:

1. The sensor space noise is scaled to equalize the mean sensor-space variance of broadband noise and brain activity.

2. The brain activity and noise are mixed to achieve the desired level of sensor space noise (denoted by \(\gamma\) below):

\[y = \sqrt{1 - \gamma} \cdot y_{brain} + \sqrt{\gamma} \cdot y_{noise}\]

If the sensor noise is independent from projected brain activity, the following relationship will hold for the total sensor space power:

\[P_{total} = (1 - \gamma) \cdot P_{brain} + \gamma \cdot P_{noise}\]

to_stc()[source]#

Obtain an stc object that contains data from all sources in the configuration.

Returns:

stcSourceEstimate: The resulting stc object that contains data from all sources.

Examples using `meegsim.configuration.SourceConfiguration`#

Using functions from other packages

Minimal example

Random state and reproducibility

Plotting the configuration

Adjustment of global SNR

Adjustment of local SNR

Adding sensor space noise

meegsim.configuration.SourceConfiguration#

Examples using meegsim.configuration.SourceConfiguration#

Examples using `meegsim.configuration.SourceConfiguration`#